thriftFind

The production and overconsumption of fast fashion has detrimental effects worldwide. We aim to reduce the need for it, in order to reduce its demand.

From its very source, fast fashion has huge environmental and human costs. Clothing production takes place in developing countries where labour costs are low, so low that factory workers are often forced to sew for 14 hours a day, 7 days a week for as little as 20 cents an hour. Meanwhile, the conditions of the factories often pose huge health risks to the workers inside; factories have been known to collapse because of poor infrastructure, killing the staff. One tragic example of this was the Rana Plaza building collapse in Dhaka in 2013, where over 1000 people died. On top of this, the vast amounts of textiles required for rapid clothing production have innumerable impacts on the environment. Cotton, as one example, requires great volumes of water to grow, resulting in local droughts and vast quantities of pesticides which run off into local water sources, reducing biodiversity. Pesticides are also thought to increase cancer in the farmers and the local people exposed long-term to them. 

In the UK clothing is worn an average of 7 times before it is discarded or donated, displaying how the rise of fast fashion has made clothing far too dispensable. Moreover, as fast fashion companies constantly change stock to keep up with new trends, tonnes of deadstock clothing they're unable to sell is also discarded. In the UK as little as 10% of donated clothing is then resold.

The UK and US both export much of their wasted clothing to countries in the global south such as Ghana and Chile. With the rise of fast fashion, the quality is often too poor to be resold and so builds up as waste in the local environment. A particularly poignant example of this is in Chile where 742-acres of used clothing have been dumped in the Atacama desert. The giant pile can now be seen from space.

Fast fashion affects billions of lives globally and is responsible for as much as 10% of global carbon emissions. It is exploiting natural resources and destroying the environment. 

We are developing a personal shopping assistant called Nova that uses AI to search across multiple secondhand and sustainable clothing sites at once and find the best pieces suited to the user, in seconds. It uses the large language model GPT-4 to effectively communicate with the user in a chat-style interface and utilises memory learning to understand their preferences over time. The personal shopping assistant will save the user both time and money, encouraging them to shop sustainably more often. 

Demo link: https://www.loom.com/share/cc0...

thriftFind's personal shopping assistant 'Nova' is designed for environmentally and ethically conscious shoppers who are trying to avoid fast fashion, but are struggling to find sustainable alternatives. These people tend to be young, between 16-30, and majority are female. 

To shop sustainably at the moment you either have to spend vast amounts on sustainable brands like Reformation or you must scroll endlessly on secondhand e-commerce sites like eBay; to find anything good in charity and thrift shops also requires lots of energy and time. These shoppers need a faster, easier and cheaper way to shop sustainably. 

We aim to make finding preloved and sustainable fashion simpler for these individuals. thriftFind's personal shopping assistant 'Nova' will search for them, avoiding the need to endlessly scroll and will be tailored to their preferences, meaning every item she finds will be relevant to them. Nova will also allow them to easily price compare between similar items, allowing them to save money on their purchases. 

As the team lead, Francesca Roberts, I have tried to shop ethically and sustainably most of my life and so have realised firsthand how difficult it is to do so, especially when you have little disposable income. Many around me have a similar experience in struggling to shop sustainably and have voiced that they need a better solution. I have done many surveys and interviews with people in this position to try and better understand the reasons they turn to fast fashion. 

A close friend of mine also works with the OR Foundation, a company dedicated to reducing the effects of the textile waste dumped in Ghana. She educated me on the impacts of fast fashion in places such as Ghana and inspired me to build a solution within the UK that could reduce the waste we dump there. We can work together in the future to build solutions for those affected by fast fashion in countries such as Ghana, using the money thriftFind makes. 

Vin has the perfect skill set to complement mine, being one of the best hackers in the world. He joined the team as the software excited him and he felt it was solving a problem he himself had. 

We initially created a video demo-ing the concept and released in on thriftFind's website. Within just one week of posting it, over 30 people submitted their email in a form on the site confirming they want access to the software when it is released.

We have now built a working prototype that can find listings from eBay and are currently testing it on a few volunteers to gain feedback. We have also started connecting with small sustainable clothing companies to discuss partnering in the future.

We will be releasing the first live version on the software by the end of the month. 

Our main reason for applying to Solve is to become part of an incredible network of people who can inspire, help, and push the idea from the prototype stage to a fully scaled business. Gaining exposure in the media and having the opportunity to learn invaluable skills through the workshops would also be incredible. Solve will be an excellent place for us to maximise our impact and will work as a catalyst to achieve our sustainable fashion goal.

The main differences between Nova and the other software currently available for shopping sustainably online are that it searches for the user, can give listings from multiple sites and has a chat-style interface. Currently are many different e-commerce sites selling preloved fashion, but there is no way to search all of them at once. Moreover, they all require endlessly scrolling to find anything good. Nova allows for easy comparison between listings from different sites and saves the user time. The chat-style interface makes shopping secondhand/sustainably fun, as Nova will be like a friend the user can talk to for advice/tips.

AI is barely integrated into any of the current sustainable shopping tools. Not only this but despite AI being a buzzword in recent years, few companies working to solve global issues are using it and many people are still fearful of what is might do. We believe thriftFind can lead the way in showing that AI can be used for good and can be used to solve some of the most pressing issues of our time. 

Finally, thriftFind will donate 50% of the profits produced to combat the issues inflicted on the planet by fast fashion. Few fashion companies are doing this, and so we believe this will set an example in the industry of the importance of prioritising positive impact over GDP. We hope others would then follow suit. 

The initial impact of thriftFind will be that more people have access to and more people can afford sustainable clothing. thriftFind will also help more clothing be recycled. On the site there will also be the opportunity for people to learn more about the impacts of fast fashion.

The long-term impacts of the software will be numerous, however. Increased access to sustainable clothing and more education on the impacts of fast fashion means there could be a reduced demand for fast fashion. This is turn reduces both the demand for exploitative labour and the demand for textile production that pollutes the Earth. The increased circulation of clothing will also help reduce textile waste, which in turn will reduce the amount of clothing the UK exports to countries in the global South and the amount of clothing which then pollutes the environment there. 

Finally, thriftFind will use 50% of its profits to combat the issues inflicted on people and the planet by fast fashion. This will have many positive impacts on the communities most affected. This may involve clearing textile waste, building better infrastructure in garment factories, protecting farmers' and workers' rights and more. 

We will measure our impact by the number of users and the number of items bought through the site, as this will give us a clear idea of how many people we are making sustainable fashion accessible to and how many pieces of clothing are being reused as a result of our software. We can also calculate the amount of CO2 saved as a result of that clothing being reused. 

At the moment we plan to calculate the impact of our company in terms of the amount of CO2 it produces. As a sustainable company, we want to be aware of and transparent about this, so that we can work towards reducing its value. For example, we are looking at using an environmentally conscious data server that offsets its CO2 consumption. 

For our present prototype, we have chosen a tech stack that emphasizes reliability and scalability for the future.

1. Our frontend utilizes React.js and Tailwind CSS, 

2. The backend leverages the Python-based framework Flask, MongoDB as our database, and Auth0 for user authentication. 

3. Regarding the language model, we are currently integrating OpenAI's GPT model API. In the near future, we plan to experiment with various other models, including Llama 3.

We believe that technology is just a tool, and we aim to extend our systems in the near future. We are consistently working with Python libraries to improve our models and provide better predictions.

Currently we have two members working part-time on the solution. Francesca, the Team leader, in charge of branding, outreach and decision making and Vin, the software developer, in charge of building the prototype.

We officially started working four months back, and we have spent the initial two months focused on researching the problem and interviewing the stakeholders. Once we got in the Youth STEM 2030 program we started with developing the initial prototype while Vin joined the project around a month back taking over the responsibility of developing a full-stack prototype ready for production.

Our team, despite only consisting of two people, is already diverse. The team lead, Francesca is a female founder and is from the UK, while Vin, is male and from India. We are both incredibly passionate about equity and ensuring our company remains diverse as it grows.

We provide value both to individual users, as they will be able to save time and money when shopping for sustainable fashion, and to companies/people selling sustainable fashion, as the Nova will increase traffic to their sites and help them make more sales. Small sustainable clothing companies would greatly benefit from this in gaining exposure. Sellers on re-commerce marketplaces will also benefit as they will make more money from their old clothing. 

We will initially rely on funding to build the software, such as from business competitions like the Summer Startup accelerator in Edinburgh. When people start using it we will charge a commission fee to the companies/sellers partnered with Nova, based on how many more sales they make as a result of thriftFind. This will be set up via affiliate programs. In the future, we may also charge the user a smaller finder fee of around 3% on top of any item Nova finds for them. This will not exist at the moment, however, as we do not want to deter initial users.